FAUNA of AUSTRALIA 36. BIOGEOGRAPHY AND PHYLOGENY OF EUTHERIA G.M. MCKAY, J.H. CALABY & L.S. HALL 1 36. BIOGEOGRAPHY AND PHYLOGENY OF EUTHERIA 2 36. BIOGEOGRAPHY AND PHYLOGENY OF EUTHERIA INTRODUCTION In considering the biogeography and phylogeny of the eutherian mammals in Australia, a distinction must be made between those groups which invaded the continent prior to the 18th Century and those that have been deliberately or inadvertently introduced since then. For the latter group, the worldwide distribution and fossil history is essentially irrelevant except that many species have a long association with humans as either domesticated animals or commensals. In this chapter, we will address the more general questions of the origins and affinities of the native Australian eutherians. For details of the biogeography, fossil history and affinities of the introduced mammals, the reader should refer to the family accounts (Chapters 44–46, 54, 55, 58–62). Any examination of the phylogeny of a group must consider two main lines of evidence: the fossil record and the study of living forms. Morphological studies can be made on both living and fossil forms and such studies, particularly of skeletal and dental characters, have traditionally formed the basis for phylogenetic reconstruction in the Mammalia. In recent years there has been an increase in biochemical and molecular studies on living species which, with a very few exceptions, cannot be compared to extinct species. To date, only a relatively few species of living mammals have been studied adequately using molecular techniques, but attempts are being made to reconcile the biochemical data with the traditional morphological studies (Shoshani 1986; McKenna 1987). Within the Eutheria, with a few exceptions such as the rodents, the taxonomic arrangements and the inferred phylogenetic relationships of families within orders have remained relatively stable since Simpson’s (1945) treatment. At the intrafamilial level, however, new techniques have led to frequent rearrangements based on the detection of substantial numbers of new biological species and a rearrangement of the content of many genera. At the supraordinal level, molecular data combined with some new anatomical studies has led to a number of changes in our understanding of the relationships between orders. We have attempted to follow the most recent revisions at the family level and below. At the supraordinal level, however, we prefer to follow the widely used classification of McKenna (1975). FOSSIL HISTORY Order Chiroptera The earliest known fossil bats are found in Early to Middle Eocene deposits of Europe and western North America. Six genera belonging to three families and a further three of interdeterminate affinity have been described (Hill & Smith 1984). The best known of these Eocene fossils, Icaronycteris index from Wyoming, has a number of distinctive microchiropteran characters, but has a claw on manal digit 2, a character which in living forms occurs only in the Megachiroptera. These Eocene bats, together with Archaeopteropus from the Oligocene of Europe which Hill & Smith (1984) considered a pteropodid, are grouped together into the superfamily Palaeochiropterygoidea within the suborder Eochiroptera (Hand 1984). According to Hand (1984), the dental characteristics of Icaronycteris are consistent with it being close to the ancestry of vespertilionoid microchiropterans, but the earliest known members of the other living microchiropteran superfamilies have specialised dentitions that indicate that modifications to the teeth may have preceded the early radiation of this suborder. Novacek (1987) described the advanced auditory specialisations 3 36. BIOGEOGRAPHY AND PHYLOGENY OF EUTHERIA of Icaronycteris and Palaeochiropteryx and considered them to be true microchiropterans, thus placing the validity of the suborder Eochiroptera in doubt. Rhinolophoids and vespertilionoids are first found in the Middle Eocene and the earliest emballonuroid occurs in Late Eocene or Early Oligocene deposits, all of which are from Europe. Following their earliest appearance in the Eocene, fossil microchiropterans of all three superfamilies appear sporadically in the Tertiary deposits of Europe and Africa with occasional appearances of some families in the Americas and Asia. Only the superfamily Rhinolophoidea is represented in the Tertiary of Australia (Hand 1984). Macroderma godthelpi and another unnamed magadermatid occur in Miocene deposits from the Riversleigh deposits of northern Queensland along with a hipposiderid (Hipposideros (Brachipposideros) nooraleebus Sigé, Hand & Archer 1982) and an undescribed rhinolophid (Hand 1984, 1985). The ancestry of the Megachiroptera is problematical. The only known Tertiary fossil is Propotto leakeyi, a pteropodid from the Early Miocene of Africa. Although most modern systematists have considered the Megachiroptera and Microchiroptera to be a monophyletic group sharing a common winged ancestor, there is considerable evidence to suggest a polyphyletic origin (Smith 1977; Pettigrew 1986; Pettigrew & Jamieson 1987). Pleistocene bats from Australia, which were listed by Archer, Clayton & Hand (1984), are all referable to modern genera and species. Order Primates For a description of the fossil record of hominids in Australia the reader is referred to Chapter 44. Order Rodentia Although rodent fossils are known as early as the Paleocene genus Paramys of North America, the earliest fossil referable to the family Muridae is Antemus from the mid Miocene Siwalik deposits of Pakistan (Carleton & Musser 1984). The earliest known Australian murids are two undescribed forms from the early Pliocene Bluff Downs deposits in northern Queensland and the early to mid Pliocene Chinchilla deposit of south-eastern Queensland (Archer et al. 1984). The only other fossil murids in Australia are a large number of Pleistocene specimens all of which are referable to extant genera and species (Archer et al. 1984). Order Carnivora The oldest known pinniped fossil, Enaliarctos from the Early Miocene of western North America, was considered by King (1983b) to be ancestral to the Otariidae which first appear in the mid-Miocene. King (1983b) described a Pliocene to Pleistocene radiation of otariids which involves the migration of two ancestral stocks to the Southern Hemisphere. The Southern fur seals are all placed in the genus Arctocephalus, but there has been further radiation in the sea lions. The only Australian fossil is a Pleistocene record of the modern Australian sea lion, Noephoca cinerea, from Queenscliff, Victoria (Archer et al. 1984). Phocid seals are considered to have evolved about the beginning of the Miocene in the North Atlantic rather than in the Pacific and to have radiated early into two lineages, the Phocinae and the Monachinae (King 1983b). Only the latter group invaded the southern oceans and, although the timing and the number of migrations involved is unknown, must have occurred early, as undescribed monachines are recorded from Late Miocene, Early Pliocene and early Pleistocene deposits in Victoria (Fordyce & Flannery 1983; Archer et al. 1984). 4 36. BIOGEOGRAPHY AND PHYLOGENY OF EUTHERIA The Australian fossil record of the Dingo (Canidae) is described in Chapter 54. Order Sirenia The only purported Australian fossil sirenian reported in the literature appears to be indeterminate (Fordyce 1982b). There is no evidence for the time of appearance of sirenians in Australia. Order Cetacea The fossil history of whales in Australia has been reviewed by Fordyce (1982b, 1984, 1985). No Australian records exist of the suborder Archaeoceti, the primitive toothed whales first found in Middle Eocene Tethyan deposits. These earliest whales, which are placed in the family Protocetidae, appear not to have had a southern distribution. The second family of archaeocetes, the Basilosauridae, occurs in Late Eocene deposits on Seymour Island and New Zealand, but not in Australia. The modern suborder Odontoceti appears first in the fossil record in the Late Oligocene and two early radiations (Squalodontidae and Rhabdosteidea) are represented in Australian deposits ranging from Late Oligocene to Middle Miocene. A further early family, the Kentriodontidae or ancestral dolphins, occurs in Late Oligocene New Zealand deposits. The first modern families to appear are the Ziphiidae and Physeteridae in Early Miocene South American deposits and are found later in the Miocene in Australia. Delphinidae occur in Late Miocene to Early Pliocene Australian deposits, but other modern families are not recorded from Australia. Whales of the suborder Mysticeti appear first in Late Oligocene deposits. Mammalodon is a Late Oligocene toothed mysticete of uncertain affinities. Other early toothed mysticetes are grouped into the family Cetotheriidae which is found in Late Oligocene deposits in New Zealand, but do not occur in Australia until the early Miocene. The family Balaenidae, which first appears in the Early Miocene of South America, occurs in Late Miocene to Early Pliocene deposits in Australia along with the first representatives of the Balaenopteridae. As with the odontocetes, many of the specimens of fossil mysticetes from modern families are referable to living genera. PHYLOGENY Eutherian Origins Eutherians and metatherians are presumed to have evolved from a common therian ancestral stock in the mid to Late Cretaceous with the earliest eutherian fossils found in both Asia and western North America (Keilan-Jaworowska, Brown & Lillegraven